The European Central Bank’s quantitative easing, first announced in January last year, has – as its most visible effect – reduced the euro’s value against the dollar. It has had little impact on headline inflation, largely driven by oil price movements. But the ECB’s programme of purchasing €1tn worth of bonds has had some unwelcome side-effects, by increasing intra-euro area imbalances among national central banks, the so-called Target-2 balances.

The beginning of QE coincided with a major Target-2 trend reversal. In January 2015, the Bundesbank experienced a €50bn increase in Target-2 claims, the biggest single monthly rise since February 2012. Since then, the German central bank’s Target-2 balances have increased constantly, reaching €600bn by January 2016. The Target-2 liabilities of Greece, Italy, Ireland, Portugal and Spain have risen correspondingly, accounting for a combined increase of €183bn Target-2 liabilities until the end of January. This is a sizable number compared to €600bn in total QE purchases up to this point.

The Target-2 balances have attracted attention since they started to rise markedly in 2007-08. The Institute of Empirical Economic Research at Osnabrück University has been monitoring them, based on monthly NCB data. In September 2015, the ECB started to release these data as monthly averages. The Osnabrück web page is now being updated using both sources.

There are three possible explanations for the Target-2 resurgence. Bundesbank President Jens Weidmann has pointed out that, under QE, NCBs purchase bonds from foreign as well as domestic holders. For instance, if the Banca d’Italia purchases an Italian bond from a German resident and this investor – after selling the bond – keeps the liquidity in Germany, this creates Target-2 balances. Weidmann concluded that this changes the interpretation of Target-2 balances away from a signal of financial market tension towards a direct effect of QE.

Another reason could be a new wave of capital flight. If the Banca d’Italia purchases the bond domestically, and the previous bondholder uses the money to buy a house in Berlin, this will drive up Target-2 balances. The condition is that the previous house owner keeps the liquidity in the country and does not balance the transaction, for instance, by buying real estate in Italy.

Additionally, NCBs may be buying bonds outside the euro area. If the assets are purchased in London, New York or elsewhere, the Target-2 values will also go up if the seller wants to convert the proceeds into assets in Germany. In all three cases, an NCB such as the Banca d’Italia is building up liabilities vis-à-vis the euro system, while Germany is accumulating claims. These claims and liabilities are equivalent to an indirect loan among central banks, similar to an international swap-line.

For policy-makers, it is worrisome that a substantial amount of liquidity created through QE is not staying in the jurisdiction of the purchasing NCB. Countries with liquidity problems experience a smaller benefit from QE, while countries with excess liquidity are flooded further.

From the point of view of taxpayers in creditor countries, the QE programme implies a higher cost from a potential euro break-up or the exit of individual countries. When Greece was on the brink of euro exit in summer 2015, potential Target-2 losses were certainly on policy-makers’ minds.

Increased Target-2 balances highlight an important aspect of euro area open market operations. Any central bank’s purchases of bonds expand its balance sheet – both assets and liabilities. But in Europe, with free movement of capital, the liabilities can move quickly between several countries’ NCBs – and can easily end up in Germany, the largest creditor. The risk-sharing agreement that was reached before QE started, as a key Bundesbank condition for participation in the arrangement, focused solely on the asset side – and therefore missed a major element of additional risk.

Coinciding with renewed worries in the last few days about possible Greek debt restructuring, the surge in Target-2 imbalances raises fresh doubts about the sustainability of the current structures behind monetary union and highlights the need for reforms.

Frank Westermann is Professor of Economics at Osnabrück University.